[R-SIG-Win] Default mtune for R windows
Uwe Ligges
ligges at statistik.tu-dortmund.de
Wed Feb 1 16:55:53 CET 2017
On 01.02.2017 08:32, Avraham Adler wrote:
> https://bugs.r-project.org/bugzilla/show_bug.cgi?id=16536
>
> I closed it after it was open over a year.
I just see this,
I can try to make a CRAN run with that option enabled and see how much
dropout we will find.
In the past, going to different opimization levels was never as easy as
expected.
In the above mention feature request, you cite that in 'GCC 4.9, this
option has been "retuned for better support of Intel core and AMD
Bulldozer architectures. Performance of AMD K7, K8, Intel Pentium-M, and
Pentium4 based CPUs is no longer considered important for generic." [3].'
Which is interesting, as we still have lots of AMD K8 machines running
at our department. If the generated code still works with a minor
penalty on those machines, we may consider it, but let me trigger a test
run in the next one or two weeks.
Best,
Uwe Ligges
>
> Avi
>
> On Wed, Feb 1, 2017 at 1:08 AM, Jeroen Ooms <jeroen.ooms at stat.ucla.edu>
> wrote:
>
>> On Tue, Jan 31, 2017 at 8:31 PM, Avraham Adler <avraham.adler at gmail.com>
>> wrote:
>>> I put in a bug request for this over a year ago and it went nowhere.
>>
>> Where did you post this exactly? If we open a topic with a working
>> patch on bugs.r-project.org and reference the canonical gcc source
>> that you are citing below, perhaps it will get merged.
>>
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-windows mailing list
> R-SIG-windows at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-windows
>
More information about the R-SIG-windows
mailing list